Speech Power Research Articles

Phase-Sensitive Joint Learning Algorithms for Deep Learning-Based Speech Enhancement

This letter presents a phase-sensitive joint learning algorithm for single-channel speech enhancement. Although a deep learning framework that estimates the time-frequency (T-F) domain ideal ratio masks demonstrates a strong performance, it is limited in the sense that the enhancement process is performed only in the magnitude domain, while the phase spectra remain unchanged. Thus, recent studies have been conducted to involve phase spectra in speech enhancement systems. A phase-sensitive mask (PSM) is a T-F mask that implicitly represents phase-related information. However, since the PSM has an unbounded value, the networks are trained to target its truncated values rather than directly estimating it. To effectively train the PSM, we first approximate it to have a bounded dynamic range under the assumption that speech and noise are uncorrelated. We then propose a joint learning algorithm that trains the approximated value through its parameterized variables in order to minimize the inevitable error caused by the truncation process. Specifically, we design a network that explicitly targets three parameterized variables: 1) speech magnitude spectra; 2) noise magnitude spectra; and 3) phase difference of clean to noisy spectra. To further improve the performance, we also investigate how the dynamic range of magnitude spectra controlled by a warping function affects the final performance in joint learning algorithms. Finally, we examined how the proposed additional constraint that preserves the sum of the estimated speech and noise power spectra affects the overall system performance. The experimental results show that the proposed learning algorithm outperforms the conventional learning algorithm with the truncated phase-sensitive approximation.

Romanticism, Self-Canonization, and the Business of Poetry. Michael Gamer. Cambridge: Cambridge University Press, 2017. Pp. vii+307.

<i>Romanticism, Self-Canonization, and the Business of Poetry</i>. Michael Gamer. Cambridge: Cambridge University Press, 2017. Pp. vii+307.

Wind noise reduction for a closely spaced microphone array in a car environment

This work studies a wind noise reduction approach for communication applications in a car environment. An endfire array consisting of two microphones is considered as a substitute for an ordinary cardioid microphone capsule of the same size. Using the decomposition of the multichannel Wiener filter (MWF), a suitable beamformer and a single-channel post filter are derived. Due to the known array geometry and the location of the speech source, assumptions about the signal properties can be made to simplify the MWF beamformer and to estimate the speech and noise power spectral densities required for the post filter. Even for closely spaced microphones, the different signal properties at the microphones can be exploited to achieve a significant reduction of wind noise. The proposed beamformer approach results in an improved speech signal regarding the signal-to-noise-ratio and keeps the linear speech distortion low. The derived post filter shows equal performance compared to known approaches but reduces the effort for noise estimation.

J. M. Coetzee’s Elizabeth Costello: From Anthropocentrism towards Biocentrism?

In this article, I argue that J. M. Coetzee’s multigeneric text, Elizabeth Costello (2003. London: Vintage Books), may be read as developing a voice for the new millennium. Coetzee can be seen to be approaching, through the figure of Elizabeth Costello, a new dimension in human/animal relations. This is one that refuses to subordinate animals to the power of human speech and writing. Moving equally away from anthropocentrism, and from notions of the ‘anthropological machine’, the dimension could be described as biocentrism, or as an approach to ecological ethics. Central to my argument is Jacques Derrida’s seminal essay ‘The Animal that Therefore I Am’ ([1997] 2002. Critical Inquiry 28 (2): 369–418) which I suggest is an intertext for Elizabeth Costello because its thinking is so congruent with Coetzee’s. I place my analysis of the Derrida text between my first and final sections on Elizabeth Costello, as the Derrida text throws light both on Costello’s initial anthropocentrism and, through Derrida’s concept of the animot, on the biocentrism towards which I argue Costello is moving.

Towards mobile gaze-directed beamforming: a novel neuro-technology for hearing loss.

Contemporary hearing aids are markedlylimited in their most important role: improving speech perception in dynamic "cocktail party" environments with multiple, competing talkers. Here we describe an open-source, mobile assistive hearing platform entitled "Cochlearity" which uses eye gaze to guide an acoustic beamformer, so a listener will hear best wherever they look. Cochlearity runs on Android and its eight-channel microphone array can be worn comfortably on the head, e.g. mounted on eyeglasses. In this preliminary report, we examine the efficacy of both a static (delay-and-sum) and an adaptive (MVDR) beamformer in the task of separating an "attended" voice from an "unattended" voice in a two-talker scenario. We show that the different beamformers have the potential to complement each other to improve target speech SNR (signal to noise ratio), across the range of speech power, with tolerably low latency.

The Ethics Resistance

Legal ethics complaints have been filed against several of the high-ranking lawyers in the Trump Administration. While individual complaints have caught the public eye, the collective movement to use legal ethics to resist Trumpism has escaped attention. This is not altogether surprising: legal ethics rules have not historically been an attractive tool for political change. Perhaps, desperate times have called for desperate measures. This “Ethics Resistance” will face significant opposing forces. Legal ethics complaints seldom result in punishment. Further, the agencies that are asked to investigate these high-profile and controversial matters will not be eager to leave their comfort zone; they are accustomed to complaints from clients who are unhappy with their lawyers over straightforward matters like unreturned phone calls or high fees. There will also be loud dissenting voices from those who will see the movement as the weaponization of a tool designed for the modest task of lawyer self-governance. Finally, the complaints will have to navigate between powerful constitutional protections of lawyer speech and federal power. But we should not dismiss this movement simply because it is unusual or challenging. The wisdom of the Ethics Resistance can only be judged after we understand its distinctive qualities and consider how they further or hinder our legal, institutional, and pragmatic interests. In this Article, I undertake that analysis and conclude that the movement has the capacity to be legally permissible, institutionally sound, and prudent. I further offer a list of best practices for future complaints.

The Securitization of Migrant Workers in Sabah, Malaysia

In the past decades, the Malaysia’s economy, particularly in Sabah, faced high dependence on migrant workers, predominantly Indonesian and Filipino workers. This over-reliance on migrant workers made the ruling elites in the country fear that their dominant presence would undermine the government’s policy to move Malaysia from a labor intensive to an automation in order to achieve the status of a developed nation, as well as a policy to prioritize locals over foreigners in all economic sectors. In order to implement this policy and to break the economic sectors from continuing to rely on the foreigners, the ruling elites have continuously associated migrant workers, especially illegal laborers, as a security threat that needs to be flushed out. This paper utilizes the Copenhagen School framework of securitization to explain why the securitizing actors, namely the politicians and the ruling elites, continue to frame Indonesian and the Filipino workers in Sabah as security issue. Two case studies are presented to examine the securitization of migrant workers in Sabah: first, “All-out war against illegals” and Ops Nyah II, 2002–03 and second, Ops Nasihat, 2004–05. However, this paper argues that making illegal migrant workers as security issue and form a sound policy to control immigrants cannot solely rely on the speech act and power of persuasion, as central in the theory because other condition factors, namely the domestic politics of the state, the federal-state relations, nature of state economy, and international pressures, are also significant in explaining why the government has never been consistent in its policy on migrant workers in Sabah.

Plato’s Gorgias and the Power of Λόγος

Abstract: Recent interpretations of the opening exchange between Socrates and Gorgias in Plato’s Gorgias have tended to focus on the question of whether the rhetorician believes that justice and injustice are the subject matter of rhetoric. This paper argues that Gorgias is interested in the just and the unjust only insofar as being a persuasive speaker on these topics is a prerequisite for the successful exercise of power in the political domain. Although Gorgias’ orientation by power and overestimation of the power of speech have been noted by many interpreters, my intention is to clarify Gorgias’ claims for the power of rhetoric and explain the significance of these claims for one important aspect of Plato’s broader condemnation of sophistry.

بەرجەستەكرنا هـؤنـەرىَ جـوانـكـاريـێ د مـەتـەلا فـؤلكلۆريـا كـورديـدا

Riddle is a significant literary art of nation, setting up the form of parse and having a great power of expression. It incorporates a partial of fine a arts that becomes a bridge of the super world of aesthetic art to show the beauty in the yarn weaved in the folklore of riddle. Fully colorful, whether in speech or meaning to show beauty in the demonstration of riddle. the elegance and decoration of riddle put it in an effective literary format and this gives the speech power full of aesthetic semantic expression by beautifying vocabulary. The existence of riddles in literature shows salient creativity of the nation and making its level of semantic expression very significant.

A new time-frequency binary mask estimation method based on convex optimization of speech power

In conventional computational auditory scene analysis, the segment segregation and pitch estimation are necessary. However, it is hard to obtain an accurate pitch contour of clean speech for segregating the segments in low signal-to-noise ratio. This often leads to an inaccurate estimation of binary mask and produces artifacts and temporal discontinuity in the enhanced speech. To overcome this problem, we propose in this paper a new estimation method for binary mask based on convex optimization of speech power. In this proposed method, the segment segregation and pitch estimation are excluded and only the speech power of each Gammatone channel is considered as a key cue used for labeling the binary masks. Considering the cross-correlation between the power spectra of noisy speech and noise in each channel, the objective function of speech power is built, and the speech power is solved by the gradient descent method. Accordingly, the time-frequency units of speech and noise are labeled by computing a decision factor derived from the powers of noisy speech, estimated speech and the pre-estimated noise. The erroneous local masks are refined by time-frequency unit smoothing. The objective measurements including segmental signal-to-noise ratio, HIT-False Alarm rate, the percentage of energy loss and the percentage of noise residue and the additional subjective listening test demonstrate the effectiveness of the proposed method.

No Other Gods

<i>No Other Gods</i>

Samuel Taylor Coleridge’s Willing Belief in the Logos of Shakespeare and “The Rime of the Ancient Mariner”

&#x0D; &#x0D; &#x0D; &#x0D; The following piece is the last we have of Professor Géza Kállay. He could not revise, let alone edit his paper before its publication. Although he had expressed his doubts concerning its “merits” after the conference, and considered it rather a “curiosity”—a curious digression from his ongoing research on Shakespeare and philosophy—the editors wish to pay homage to him by conveying his words as faithfully as they can to transmit, however imperfectly, what Géza calls, in his paper, “voiced animation.” That is, the “heated passion” with which he—like Coleridge or the actors impersonating Shakespeare’s characters—used to “animate the ‘cold,’ arbitrary and con- ventional symbols [...] of everyday language” in lecture theatres, in seminar and conference rooms. Géza had a “strange power of speech” that used to mesmerise his students and colleagues alike.&#x0D; &#x0D; &#x0D; &#x0D;

Daya Tutur Metafora Lirik Lagu Populer (Kajian Pragmatik)

Communicating can be done with any media as needed and function, one of them communication through song lyrics. Song lyrics are created to have a specific purpose in addition to being an aesthetic creative means by using figurative language in the form of metaphors can also be a means of speaking with the audience. Speech done is empowered speechless illocution and perlokusi. Ilocution is the act of doing something. The act of illocution is a speech act that contains the purpose and function or the power of speech. The question posed related to the act of illocution is For what the speech was done.&#x0D; While Perlokusi speech that has the effect or influence power. The effect or power of the speech may be caused by the speaker intentionally or unintentionally. The study used is a pragmatic study. The data is taken from the lyrics of Indonesia's popular metaphorical song.

Whose Speech Is It Anyway? Ownership, Authorship, and the Redfern Address

Whose Speech Is It Anyway? Ownership, Authorship, and the Redfern Address

Perceptual Implications of Level- and Frequency-Specific Deviations from Hearing Aid Prescription in Children.

The purpose of providing amplification for children with hearing loss is to make speech audible across a range of frequencies and intensities. Children with hearing aids (HAs) that closely approximate prescriptive targets have better audibility than peers with HA output below prescriptive targets. Poor aided audibility puts children with hearing loss at risk for delays in communication, social, and academic development. The goals of this study were to determine how well HAs match prescriptive targets across ranges of frequency and intensity of speech and to determine how level- and frequency-dependent deviations from prescriptive target affect speech recognition in quiet and in background noise. One-hundred sixty-six children with permanent mild to severe hearing loss who were between 6 months and 8 years of age and who wore HAs participated in the study. Hearing aid verification and speech recognition data were collected as part of a longitudinal study of communication development in children with HAs. Hearing aid output at levels of soft and average speech and maximum power output were compared with each child's prescriptive targets. The deviations from prescriptive target were quantified based on the root-mean-square (RMS) error and absolute deviation from target for octave frequencies. Children were classified into groups based on the number of level-dependent deviations from prescriptive target. Frequency-specific deviations from prescriptive target and sensation levels (SLs) were used to estimate the proximity of fittings across the frequency range. Lexical Neighborhood Test (LNT) word recognition in quiet and Computer-Assisted Speech Perception Assessment (CASPA) phoneme recognition in noise were compared across level-dependent error groups and as a function of SL at 4 kHz. Children who had deviations from prescriptive target at all three input levels had poorer LNT word recognition in quiet than children who had fittings that matched prescriptive target within 5 dB RMS at all three input levels. Children with lower 4 kHz SLs through their HAs had poorer LNT recognition in quiet and CASPA phoneme recognition in noise than children with higher aided SLs. Children with HAs fitted to provide audibility for speech across a range of inputs and frequencies had better speech recognition outcomes than peers with HAs that were not optimally fitted to prescriptive targets.

A Trauma-Informed Approach to the Protection and Support of Witnesses in International Tribunals: Ten Guiding Principles

Drawing on the scholarly and clinical work of trauma experts, this article presents a conceptual framework to inform the provision of services for witnesses in international tribunals. In particular, the authors present the following ten principles to guide protection and support practices with witnesses: establish a caring relationship; maintain a safe environment; stabilize physiological and mental functioning; maximize predictability; maximize choices; affirm dignity and value; affirm the power of speech; promote human connection; instil hope; and foster reconnection with ordinary life. The authors discuss these principles according to three clusters, based on their relative salience at different phases of a witness's participation: pre-testimony, during testimony, and post-testimony. For each principle, the authors provide a rationale as well as some examples from the witness unit at the Special Court for Sierra Leone. They also discuss some of the challenges encountered in the effort to adhere to each principle. With modifications to ensure relevance to specific contexts and cultures, the adoption of these principles will enhance the likelihood of positive experiences, and reduce the risk of harm, among individuals who testify in international war crimes trials.

Performativity and the Persistence of the Golem Legend

This article investigates The Golem by H. Leivick and posits the 1921 play as a harbinger of the theory of performativity that has had a significant impact upon theatre and performance studies in the late twentieth and early twenty-first centuries. The golem of ancient Jewish legend was a clay figure brought to life through the incantations of learned rabbis. The traditional golem possessed many human qualities but not the uniquely human power of speech – the power that brought it to life. In Leivick’s play, however, the Golem can speak just as eloquently as his creator. This equal access to speech suggests, as do later theorists of performativity, that we are all golems of a sort, created and creating one another through the act of speech itself. This insight has crucial ethical implications for the present day.

Sparse representation‐based quasi‐clean speech construction for speech quality assessment under complex environments

A non-intrusive speech quality assessment method for complex environments was proposed. In the proposed approach, a new sparse representation-based speech reconstruction algorithm was presented to acquire the quasi-clean speech from the noisy degraded signal. Firstly, an over-complete dictionary of the clean speech power spectrum was learned by the K-singular value decomposition algorithm. Then in the sparse representation stage, the stopping residue error was adaptively achieved according to the estimated cross-correlation and the noise spectrum which was adjusted by a posteriori SNR-weighted factor, and the orthogonal matching pursuit approach was applied to reconstruct the clean speech spectrum from the noisy speech. The quasi-clean speech was considered as the reference to a modified PESQ perceptual model, and the mean opinion score of the noisy degraded speech was achieved via the distortions estimation between the quasi-clean speech and the degraded speech. Experimental results show that the proposed approach obtains a correlation coefficient of 0.925 on NOIZEUS complex environment database, which is 99% similar to the performance of the intrusive standard ITU-T PESQ, and 7.1% outperforms non-intrusive standard ITU-T P.563.

Modeling the effects of single-microphone noise-suppression

Many different algorithms have been proposed for single-microphone noise suppression. Comparing algorithms can be difficult, however, since different studies use different speech stimuli, different types of noise, and different signal-to-noise ratios (SNRs). This paper proposes the HASPI intelligibility and HASQI quality metrics as effective tools to measure and compare the performance of noise-suppression algorithms. The stimuli are sentences presented in multi-taker babble at SNRs ranging from −20 to +40dB. Six different noise-suppression algorithms are compared: Wiener, power, and magnitude spectral subtraction, a smoothed gain-vs-SNR approach that limits the maximum attenuation, binary mask processing, and auditory masked transform noise suppression. Exact restoration of the speech envelope and no processing provide reference conditions. Also investigated are the impact of differing amounts of knowledge about the noise signal and the effects of impaired versus normal hearing. In general, the algorithms that worked best for normal hearing were also most effective for a simulated moderate hearing loss. The results indicate that many algorithms provide high predicted intelligibility under ideal conditions where the speech and noise power are known for every segment within each frequency band, but that the performance degrades to that measured for no processing when the local noise power is replaced by the power averaged over all of the segments within the band. The upper bound to algorithm performance is shown to be exact restoration of the noisy speech envelope to match that of the noise-free signal.

Decision-directed speech power spectral density matrix estimation for multichannel speech enhancement

In this letter, a multichannel decision-directed approach to estimate the speech power spectral density (PSD) matrix for multichannel speech enhancement is proposed. There have been attempts to build multichannel speech enhancement filters which depend only on the speech and noise PSD matrices, for which the accurate estimate of the clean speech PSD matrix is crucial for a successful noise reduction. In contrast to the maximum likelihood estimator which has been applied conventionally, the proposed decision-directed method is capable of tracking the time-varying speech characteristics more robustly and improves the noise reduction performance under various noise environments.